On Mon, 7 Jun 2004 09:48:23 -0400, Paul Winkler wrote:
On Mon, Jun 07, 2004 at 11:21:25AM +0200, Bert Vanderbauwhede wrote:
The only thing I can conclude, is that the problem is caused by requests that somehow lock up Zope. These requests come from one or more users at the university.
Just a guess, but maybe there are some local user(s) who are infected with worms / virii that are DOSing your Zope?
That is one of the possibilities. But it doesn't have to be a virus. There is very little
I'd look through your Zope access log and see if it's only the number of requests that are a problem, or the
I was thinking of using the Apache logs to 'replay' the requests of friday, and see if I can recreate the problem. I haven't had time to do it yet.
I've seen on some of my sites that extremely large numbers of requests to non-existent pages can effectively DOS zope.
This is interresting. This can explain what happened when I got all those M$ WebDAV requests. And when I had the same problem two weeks ago, I noticed a whole bunch of requests for Apache error pages, that were proxied to Zope, but which Zope couldn't find of course.
But I've never found a solution other than waiting for the storm to subside.
Too bad. In the case of those M$ WebDAV requests, I was able to locate the users who were causing the trouble and make them stop. But then again, those WebDAV requests were easy to spot in access_log. This time, it's a bit harder.
Another guess: Do you use ZEO?
No. Bert Vanderbauwhede... -- "I laugh in the face of danger. Then I hide until it goes away." -- Alexander LaVelle Harris